The persistence of local joblessness

Local differences in US employment-population ratios and unemployment rates have persisted over many decades. Using decennial census data from 1950-2010, we investigate the reasons for this. The persistence cannot be explained by permanent differences in amenities, local demographic composition or the propensity of women to work. Population does respond strongly to differences in economic fortunes, although these movements are not large enough to eliminate shocks within a decade. Over the longer run, persistence in local joblessness is largely explained by serial correlation in the demand shocks themselves.